Structured Graph Learning for Scalable Subspace Clustering: From Single View to Multiview

• Published in: IEEE transactions on cybernetics Vol. 52; no. 9; pp. 8976 - 8986
• Main Authors: Kang, Zhao, Lin, Zhiping, Zhu, Xiaofeng, Xu, Wenbo
• Format: Journal Article
• Language: English
• Published: United States IEEE 01.09.2022; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Anchor graph; Bipartite graph; Clustering; Clustering algorithms; Clustering methods; connectivity constraint; Data points; Graph theory; Laplace equations; large-scale data; Machine learning; multiview learning; out of sample; Periodic structures; subspace clustering; Subspace methods; Training; Videos
• ISSN: 2168-2267; 2168-2275; 2168-2275
• DOI: 10.1109/TCYB.2021.3061660

Graph-based subspace clustering methods have exhibited promising performance. However, they still suffer some of these drawbacks: they encounter the expensive time overhead, they fail to explore the explicit clusters, and cannot generalize to unseen data points. In this work, we propose a scalable graph learning framework, seeking to address the above three challenges simultaneously. Specifically, it is based on the ideas of anchor points and bipartite graph. Rather than building an <inline-formula> <tex-math notation="LaTeX">n\times n </tex-math></inline-formula> graph, where <inline-formula> <tex-math notation="LaTeX">n </tex-math></inline-formula> is the number of samples, we construct a bipartite graph to depict the relationship between samples and anchor points. Meanwhile, a connectivity constraint is employed to ensure that the connected components indicate clusters directly. We further establish the connection between our method and the <inline-formula> <tex-math notation="LaTeX">K </tex-math></inline-formula>-means clustering. Moreover, a model to process multiview data is also proposed, which is linearly scaled with respect to <inline-formula> <tex-math notation="LaTeX">n </tex-math></inline-formula>. Extensive experiments demonstrate the efficiency and effectiveness of our approach with respect to many state-of-the-art clustering methods.